JP3979499B2 - System and method for automatically recognizing vehicle number of monitored vehicle using monitoring vehicle - Google Patents

Description

The present invention relates to a vehicle recognition system and method.
More specifically, by detecting the license plate of a monitored vehicle that is operating or parked using a camera installed on the surveillance vehicle, and detecting whether the vehicle is stolen, arranged, license plate can be altered, etc. The present invention relates to a system and a method for automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle so that a problem vehicle can be initially processed.

  Conventional vehicle recognition (vehicle monitoring) can be recognized through a surveillance camera installed on the main road, as we are well-known. After the police and surveillance personnel confirm the vehicle number, color, etc. with the naked eye, It was also possible to check this without using any electricity.

  However, checking the vehicle by such a method has a problem that a certain limit has to be recognized. In other words, the surveillance camera installed on the main road has the problem that the position is fixed, so that a considerable part is exposed, and not only that, but the problem vehicle is captured by the surveillance camera from that position. However, a considerable amount of time is required to connect this to the clearance of the vehicle, and the distance between the clearance personnel and the problem vehicle is large.

  Furthermore, when a person such as a road police or a police officer directly monitors, there is a problem that not only a continuous monitoring is basically impossible but also a perfect monitoring cannot be performed within the monitoring period.

  In addition, it is difficult to grasp whether or not the vehicle is a problem vehicle when instantaneously discovering the vehicle, and even if a problem vehicle is found, it is difficult to leave it as a document. there were.

  An object of the present invention is to obtain information on a monitored vehicle that is operating or parked by a monitoring vehicle equipped with a camera so that the above-described problems can be solved, and whether there is a problem with the vehicle through a database. It is an object of the present invention to provide a vehicle number automatic recognition system and method for a monitored vehicle that uses a monitoring vehicle, so that a suspicious vehicle can be initially processed.

A system according to the present invention for achieving the above-described object includes a camera that captures license plate information of a monitored vehicle that is operating or parked, and a vehicle number of the monitored vehicle based on data captured through the camera. The result data on whether or not the problem of the monitored vehicle is generated is received from the server computer that extracts and confirms whether the vehicle is a problem vehicle based on the extracted vehicle number of the monitored vehicle. Or monitoring means that informs passengers, surveillance vehicles equipped with portable communication equipment that transmits the vehicle number of the monitored vehicle extracted from the monitoring means to the server computer through the communication network, and police, scrapped vehicle business through the communication network Information on various problem vehicles stolen and arranged periodically from the computer of the processing organization that controls and manages vehicles such as offices, ward offices, etc. When the vehicle number of the monitored vehicle is received from the portable communication device provided in the monitored vehicle through the communication network, the vehicle number of the monitored vehicle is stored in the database. compares the vehicle information, check the problem occurrence whether the monitored vehicle, the result data for the problem occurrence whether the monitored vehicle, including a server computer to transmit to the mobile communication device provided in the monitoring vehicle but ,
The monitoring means includes
In order to extract an effective license plate area, after downsampling the video taken by the camera, the downsampled video is composed of video pyramids with different spatial resolutions, and the license plate color information is obtained. Use this to convert shades to gray levels, convert to binarized video by selecting the appropriate threshold value, use horizontal / vertical projection video generation, and use horizontal / vertical projection video and license plate horizontal / vertical ratio If the position and size of the license plate area are correct from the clustered video, the video is determined to be a valid license plate area and is not correct. The vehicle to be monitored by controlling the vertical and horizontal position information and zoom of the camera. Number plate information retake, extracted again a valid license plate area of
In order to extract characters from the valid license plate area, the extracted license plate area is converted to a gray level, and a horizontal histogram is applied to the license plate area converted to the gray level to divide it into an upper part and a lower part. Applying a vertical histogram to the license plate area divided into the upper part and the lower part to extract individual characters existing in the license plate area and extracting individual characters through binarization processing through appropriate threshold selection,
In order to identify the extracted individual characters, the individual characters in the extracted license plate area are divided from each individual character in the vertical / horizontal direction, and then mapped in 64 dimensions (8 × 8), and then each mapped individual character Characteristic values are extracted by normalizing character values, and individual characters are identified .

In addition, another system according to the present invention extracts a vehicle number of a monitored vehicle based on data captured through the camera and a camera that captures monitored license plate information that is operating or parked. The vehicle number of the monitored vehicle is compared with the problem vehicle information stored in the database to check whether the problem occurs in the monitored vehicle, and the result data on whether the problem occurs in the monitored vehicle is written and written by the driver. or intellectual racemase in passenger, if impossible result notice to issue occurs whether the monitored vehicle to the server computer includes a monitoring means for recording the discovery of problems vehicle, problems that are periodically updated from a server computer If the vehicle information is received, but it is possible to notify the server computer of the result of whether or not there is a problem with the monitored vehicle, it is confirmed as a problem vehicle. Surveillance vehicles equipped with portable communication devices that transmit information of surveillance vehicles to a server computer through a communication network, and computers of processing agencies that control and manage vehicles such as police, scrap car offices, ward offices, etc. through the communication network Information on various problem vehicles that were periodically stolen and arranged from is provided, a database is built, and information on problem vehicles that are periodically updated is transmitted to the portable communication device provided in each monitoring vehicle through the communication network. And a server computer that stores information on problem vehicles and confirmed monitored vehicles in a database, received from a portable communication device provided in the monitored vehicles through the communication network ,
The monitoring means includes
In order to extract an effective license plate area, after downsampling the video taken by the camera, the downsampled video is composed of video pyramids with different spatial resolutions, and the license plate color information is obtained. Use this to convert shades to gray levels, convert to binarized video by selecting the appropriate threshold value, use horizontal / vertical projection video generation, and use horizontal / vertical projection video and license plate horizontal / vertical ratio If the position and size of the license plate area are correct from the clustered video, the video is determined to be a valid license plate area and is not correct. The vehicle to be monitored by controlling the vertical and horizontal position information and zoom of the camera. Number plate information retake, extracted again a valid license plate area of
In order to extract characters from the valid license plate area, the extracted license plate area is converted to a gray level, and a horizontal histogram is applied to the license plate area converted to the gray level to divide it into an upper part and a lower part. Applying a vertical histogram to the license plate area divided into the upper part and the lower part to extract individual characters existing in the license plate area, extracting individual characters through binarization processing through appropriate threshold value selection,
In order to identify the extracted individual characters, the individual characters in the extracted license plate area are divided from each individual character in the vertical / horizontal direction, and then mapped in 64 dimensions (8 × 8), and then each mapped individual character Characteristic values are extracted by normalizing character values, and individual characters are identified .

Further, another system according to the present invention is extracted after extracting a vehicle number based on a camera that captures license plate information of a monitored vehicle that is in operation or parked and the monitored data of the monitored vehicle. Based on the vehicle number, the server computer confirms whether the vehicle is a problem vehicle. Result data for the problem occurrence of the monitored vehicle is received through a portable communication device in text and voice to the driver or passenger. A monitoring means for informing, a monitoring vehicle provided with a portable communication device that transmits data taken through the camera to a server computer through a communication network, and police, a scrap car office, a ward office, etc. through the communication network, Information on various problem vehicles that were stolen or arranged was periodically received from the computer of the processing agency that manages the database. The vehicle number of the monitored vehicle is extracted based on the photographing data received from the monitored vehicle through the communication network, and the problem vehicle stored in the database and the extracted vehicle number of the monitored vehicle To determine whether or not a problem occurs in the monitored vehicle, and transmit result data on whether or not the problem vehicle occurs in the monitored vehicle to a portable communication device provided in the monitored vehicle, Include a server computer that allows passengers to check ,
The server computer
In order to extract an effective license plate area, after downsampling the video taken by the camera, the downsampled video is composed of video pyramids with different spatial resolutions, and the license plate color information is obtained. Use this to convert shades to gray levels, convert to binarized video by selecting the appropriate threshold value, use horizontal / vertical projection video generation, and use horizontal / vertical projection video and license plate horizontal / vertical ratio If the position and size of the license plate area are correct from the clustered video, the video is determined to be a valid license plate area and is not correct. The vehicle to be monitored by controlling the vertical and horizontal position information and zoom of the camera. License plate region retake, extracted again a valid license plate area of
In order to extract characters from the valid license plate area, the extracted license plate area is converted to a gray level, and a horizontal histogram is applied to the license plate area converted to the gray level to divide it into an upper part and a lower part. Applying a vertical histogram to the license plate area divided into the upper and lower sections to extract individual characters existing in the license plate area and extracting individual characters through binarization processing by selecting appropriate threshold values ,
In order to identify the extracted individual characters, the individual characters in the extracted license plate area are divided from each individual character in the vertical / horizontal direction, and then mapped in 64 dimensions (8 × 8), and then each mapped individual character Characteristic values are extracted by normalizing character values, and individual characters are identified .

The method according to the present invention includes (1) recognizing a monitored vehicle that is operating or parked within a shooting range of a camera attached to the monitored vehicle, and includes a license plate of the monitored vehicle through the camera. Shooting, (2) extracting the vehicle number of the monitored vehicle based on the shooting data of the monitored vehicle taken by the camera, and (3) the object extracted from the monitoring means provided in the monitoring vehicle. A step of transmitting the vehicle number of the monitored vehicle to the server computer through the portable communication device; and (4) when the vehicle number of the monitored vehicle is received from the monitored vehicle by the server computer, the problem of the monitored vehicle is grasped. Depending on the availability of the database provided for the purpose, the latest problem vehicle information is input from the computer of the relevant processing organization and the database is updated. (5) comparing the information on the problem vehicle stored in the database based on the vehicle number of the monitored vehicle received from the monitored vehicle by the server computer and confirming whether or not the problem has occurred; (6) If it is confirmed that the monitored vehicle is not a problem vehicle, the server computer notifies the monitored vehicle that the monitored vehicle is not a problem vehicle, and (7) confirms that the monitored vehicle is a problem vehicle. If the server computer determines whether the monitored vehicle can be processed from the monitored vehicle, and (8) if the monitored vehicle can process the monitored vehicle in question, A step of notifying the monitored vehicle from the server computer to the monitored vehicle that the monitored vehicle is a problem vehicle so that the problematic monitored vehicle can be processed; If the problem monitored vehicle cannot be processed, the server computer reports the monitored vehicle information to the computer of the related processing engine so that the problem monitored vehicle can be processed. It is made including
The step (2) includes (2-1-1) obtaining a video for extracting a license plate region from shooting data including license plate information of a monitored vehicle taken by a camera; -2) Downsampling to lower the resolution of the video obtained from the step (2-1-1) to mitigate image quality degradation due to camera movement; (2-1-3) Suitable for license plate region extraction In order to select the resolution, the downsampled video is composed of video pyramids having different spatial resolutions, and (2-1-4) the license plate color after being composed of video pyramids having different spatial resolutions. Using information to convert the degree of hue to gray level, converting to binary video by selecting the appropriate threshold value, and projecting video in the horizontal / vertical direction And performing clustering for utilizing the horizontal / vertical ratio of adult and horizontal / vertical projection image and license plate (Aspect ratio) for extracting a region of rectangular shape, is performed (2-1-5) Clustering It is determined whether the position and size of the license plate area is correct from the captured image, the next step (2-2-1) is performed, and (2-1-6) the license plate is numbered from the clustered image. When the position and size of the area are not correct, the camera controls the camera from the control unit of the monitoring means provided in the monitoring vehicle by generating position information and zoom control signals of the upper, lower, left and right of the camera; (2-2 1) converting the license plate area extracted in the steps (2-1-1) to (2-1-6) into a gray level; and (2-2-2) converting the license plate area into a gray level. -Applying a horizontal histogram to the plate area and separating it into an upper part and a lower part; (2-2-3) Applying a horizontal histogram to the license plate area separated into an upper part and a lower part Applying and extracting individual characters existing in the license plate area; (2-2-4) extracted by applying the horizontal / vertical histogram in the above steps (2-2-2) and (2-2-3) After the individual characters are binarized through selection of appropriate threshold values, the next step (2-3) is performed, and (2-3-1) the steps (2-2-1) to (2- 2-4) a step of dividing each individual character in the license plate area in the vertical direction; and (2-3-2) dividing each individual character in the license plate region in the vertical direction and then dividing in the horizontal direction. Stage (2-3-3) after dividing each individual character in the license plate area in the vertical / horizontal direction, and then performing 64 dimensional (8 × 8) mapping, and (2-3-4) 64 dimensional After (8 × 8) mapping, the value of each mapped individual character is normalized and extracted from the feature vector .

In another method according to the present invention, (A) a monitored vehicle that is operating or parked within a shooting range of a camera attached to the monitored vehicle is recognized, and the license plate of the monitored vehicle is passed through the camera. Including the steps of: (B) extracting the vehicle number of the monitored vehicle based on the shooting data of the monitored vehicle imaged by the camera; and (C) extracting the vehicle number of the monitored vehicle; In the monitoring means provided in the monitoring vehicle, information on the latest problem vehicle provided from the server computer to the computer of the related processing engine according to the availability of the database provided for grasping the problem availability of the monitored vehicle The database is updated based on the stage of updating the database and (D) the vehicle number of the monitored vehicle extracted from the monitoring means provided in the monitored vehicle. Comparing the information on the problem vehicles being confirmed and confirming whether or not a problem has occurred; and (E) if it is confirmed that the monitored vehicle is not a problem vehicle, the monitoring means provided in the monitoring vehicle A step of outputting a character or voice notifying that the monitored vehicle is not a problem vehicle and confirming to the driver or a passenger, and (F) when the monitored vehicle is confirmed to be a problem vehicle, A step of determining whether or not the monitored vehicle can be processed by the monitoring means provided in the monitored vehicle; and (G) if the monitored vehicle having the problem can be processed in the monitored vehicle, A step of outputting a character or a voice notifying that the monitored vehicle is a problem vehicle from the monitoring means provided in the monitored vehicle so that the driver or passenger can process the monitored vehicle having a problem. And (H) in the monitoring vehicle, if processing of the monitored vehicle in question is impossible, information on the monitored vehicle is transmitted from the monitoring means provided in the monitoring vehicle to the computer of the related processing engine for portable communication. Reporting through the equipment and enabling the handling of problematic monitored vehicles ,
The step (B) includes: (B-1-1) obtaining a video for extracting a license plate region from shooting data including license plate information of a monitored vehicle taken by a camera; 1-2) A step of down-sampling to lower the resolution of the video obtained from the step (B-1-1) to mitigate the deterioration of image quality due to the movement of the camera, and (B-1-3) adapted to extraction of the license plate region. In order to select the resolution, the down-sampled video is composed of video pyramids with different spatial resolutions, and (B-1-4) is composed of video pyramids with different spatial resolutions. Using color information, the degree of hue is converted to gray level, converted to binary video by selecting an appropriate threshold value, and projected in horizontal / vertical directions. Performing clustering to extract a rectangular area using image generation and horizontal / vertical projection video and the aspect ratio of the license plate, and (B-1-5) clustering is performed. It is determined whether the position and size of the license plate area are correct from the broken video and the next stage (B-2-1) is performed, and (B-1-6) the number from the video on which clustering is performed. If the position and size of the plate area are not correct, the camera controls the camera from the control unit of the monitoring means provided in the monitoring vehicle by generating position information and zoom control signals for the camera, and (B-2 -1) and converting to said step (B-1-1) not the license plate area extracted by the (B-1-6) to the gray level, was converted to (B-2-2) gray levels A horizontal histogram is applied to the upper plate portion and the upper portion and the lower portion are separated from each other, and (B-2-3) a vertical histogram is applied to the license plate region separated into the upper portion and the lower portion by applying the horizontal histogram. To extract individual characters existing in the license plate area, and (B-2-4) extracted through the horizontal / vertical histogram application of the above steps (B-2-2) and (B-2-3). After the individual characters are binarized by selecting an appropriate threshold value, the next step (B-3) is performed, and (B-3-1) the steps (B-2-1) to (B-) are performed. 2-4) a step of dividing each individual character in the license plate area in the vertical direction; and (B-3-2) dividing each individual character in the license plate region in the vertical direction and then dividing in the horizontal direction. Stage , (B-3-3) After dividing each individual character in the license plate area in the vertical / horizontal direction, mapping 64 dimensions (8 × 8), and (B-3-4) 64 dimensions ( After 8 × 8) mapping is performed, the value of each mapped individual character is normalized to extract a feature vector .

According to still another method of the present invention, (a) a monitored vehicle that is operating or parked within a shooting range of a camera attached to the monitored vehicle is recognized, and the license plate of the monitored vehicle is passed through the camera. And (b) transmitting captured data of the monitored vehicle captured by the camera to the server computer through the communication network; and (c) receiving captured data of the monitored vehicle from the monitored vehicle at the server computer. A step of extracting the vehicle number of the monitored vehicle based on the photographing data, and (d) whether or not the database provided for grasping the problem of the monitored vehicle in the server computer is related Receiving the latest information on the problem vehicle from the computer of the processing engine and updating the database; (e) Based on the vehicle number of the monitored vehicle from the computer, comparing the problem vehicle information stored in the database to confirm whether or not the problem has occurred, and (f) confirming that the monitored vehicle is not the problem vehicle The server computer notifies the monitoring vehicle that the monitored vehicle is not a problem vehicle; and (g) if the monitored vehicle is confirmed to be a problem vehicle, the server computer Determining whether or not the monitored vehicle can be processed, and (h) in the monitored vehicle, if the problematic monitored vehicle can be processed, the server computer sends the monitored vehicle to the monitored vehicle. Reporting that the monitored vehicle is a problem vehicle and allowing the problematic monitored vehicle to be processed, and (i) the monitored vehicle having the problem in question If treatment of both is not possible, but inclusion of steps to be able to handle a monitored vehicle with by Problem issues information of the monitored vehicle to computer-related processing engine from the server computer,
The step (c) includes (c-1-1) obtaining a video for extracting a license plate region from shooting data including license plate information of a monitored vehicle taken by a camera, and (c- 1-2) A step of downsampling to lower the resolution of the video obtained from the step (c-1-1) to mitigate a deterioration in image quality due to a camera movement, and (c-1-3) adapted to extraction of a license plate region. In order to select the resolution, the downsampled video is composed of video pyramids having different spatial resolutions, and (c-1-4) the license plate color is composed of video pyramids having different spatial resolutions. Using information to convert the degree of hue to gray level, selecting the appropriate threshold value, and converting to binary image, projection in horizontal / vertical direction Clustering to extract a rectangular area using image generation and horizontal / vertical projection video and license plate horizontal / vertical ratio, and (c-1-5) from the clustered video A step of determining whether the position and size of the license plate area are correct and performing the next step (c-2-1), and (c-1-6) the position of the license plate region from the clustered video If the size is not correct, the camera controls the camera from the control unit of the monitoring means provided in the monitoring vehicle by generating position information and zoom control signals of the upper, lower, left and right of the camera, and (c-2-1) Converting the license plate area extracted in steps (c-1-1) to (c-1-6) into a gray level; and (c-2-2) converting the license plate into a gray level. Applying a horizontal histogram to the upper region and the lower portion by applying a horizontal histogram to the image region; and (c-2-3) applying a vertical histogram to the license plate region separated into the upper portion and the lower portion by applying the horizontal histogram. Applying and extracting individual characters existing in the license plate area, and (c-2-4) extracted through the horizontal / vertical histogram application of step (c-2-2) and (c-2-3) After binarizing the individual characters by selecting an appropriate threshold value, performing the next step (c-3-1), (c-3-1) the steps (c-2-1) to ( (c-3-4) vertically dividing each individual character of the license plate area extracted by (c-2-4); and (c-3-2) dividing the individual character of the license plate area vertically, and then horizontally Dividing into steps
And (c-3-3) performing a 64-dimensional (8 × 8) mapping after dividing each individual character in the license plate area in the vertical / horizontal direction, and (c-3-4) 64-dimensional After performing (8 × 8) mapping, the value of each mapped individual character is normalized to include a step of extracting a feature vector .

  At this time, in the step (a), (a1) the driver / passenger of the monitoring vehicle or the mobile (portable) camera owner is operating or parked using the direct mobile (portable) camera. The license plate information of the monitored vehicle can be photographed, and the mobile (portable) camera can be connected to the monitoring means, and then the photographing data of the monitored vehicle can be downloaded to the monitoring means.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 1 is a diagram showing a configuration of an automatic recognition system for a vehicle number of a monitored vehicle using a monitoring vehicle according to the present invention.

  As shown in the figure, a vehicle number automatic recognition system according to the present invention includes a monitored vehicle (100), a monitoring vehicle (200) equipped with automatic vehicle number recognition equipment, and a communication network (300) for information transmission. A server computer (400) that manages and stores data related to automatic recognition of vehicle numbers and performs data communication with the monitoring vehicle (200), a police, a scrap car office, a ward office, etc. An institutional computer (500).

  The monitoring vehicle (200) is monitored based on the camera (210) that captures information including information on the license plate of the monitored vehicle (100) that is operating or parked, and data that is captured through the camera (210). The vehicle number of the vehicle (100) is extracted, and the monitored vehicle (100) is checked from the server computer (400) that confirms whether the vehicle is a problem vehicle based on the extracted vehicle number of the monitored vehicle (100). ) To receive the result data for the possibility of the problem occurrence, and to notify the driver or passenger by text / voice, and the vehicle number of the monitored vehicle (100) extracted from the monitoring means (220) And a portable communication device (230) for transmitting to the server computer (400) through the communication network (300).

  At this time, as the monitoring vehicle (200), a police patrol vehicle, a vehicle in the security industry, a vehicle in operation such as a taxi is used.

  Further, when the camera (210) mounted on the monitoring vehicle (200) captures the license plate of the monitored vehicle (100), the following two methods can be used.

  First, it uses a camera photography system with automatic sensing. That is, a method is used in which a sensing sensor (not shown) attached to the camera (210) or located at another part of the monitoring vehicle (200) senses the monitored vehicle (or a license plate of the vehicle) and shoots. Is.

  Secondly, a photographing method by the driver or passenger of the monitoring vehicle (200) is used. That is, a method is used in which a driver (or a passenger) on a monitoring vehicle (200) finds the monitored vehicle (100) and presses a shooting switch, and the camera (210) that recognizes the shooting switch operation performs shooting. .

  The server computer (400) is information on various problem vehicles that are periodically stolen and arranged from the computer (500) of a processing organization that controls and manages vehicles such as police, scrap car offices, ward offices, etc. through a communication network (300). The vehicle number of the monitored vehicle (100) is compared from the portable communication device (230) provided in the specific monitored vehicle (200) through the communication network (300). 100) portable communication provided in the monitoring vehicle (200) with confirmation of whether or not the problem occurs (that is, theft, arrangement, forgery of license plate, etc.) The data is transmitted to a device (230) (for example, a TRS (Trunked Radio System) terminal, a PDA (Personal Digital Assistant), etc.).

  At this time, the server computer (400) described above is stored in the database based on the vehicle number of the monitored vehicle (100) received from the portable communication device (230) provided in the specific monitoring vehicle (200). When checking the possibility of occurrence of a problem by comparing information on various problem vehicles arranged for theft and arrangement, if the monitored vehicle (100) is confirmed as a problematic vehicle, processing by the monitored vehicle (200) is possible. If it is impossible to process the monitored vehicle (100) that is confirmed as a problem vehicle by the monitoring vehicle (200), it is necessary to determine whether the monitoring vehicle (200) Information (vehicle number, current position, etc. of the monitored vehicle) of the monitored vehicle (100) confirmed as the problem vehicle is transmitted to the computer (500) for processing.

  FIG. 2 is a view showing in more detail the internal configuration of the monitoring vehicle 200 of FIG. 1 according to the first embodiment of the present invention.

  As shown in the figure, the key operation unit (221) receives the key operation signal of the driver or passenger of the monitored vehicle (200) for recognizing the vehicle number of the monitored vehicle (100), and receives the control unit (223). ).

  The photographing data input unit (222) receives photographing data of the monitored vehicle (100) photographed through the camera (210) and outputs it to the control unit (223).

  The control unit (223) outputs the shooting data of the monitored vehicle (100) input through the shooting data input unit (222) to the vehicle number recognition unit (224) to extract the vehicle number of the monitored vehicle (100). The vehicle number of the monitored vehicle (100) extracted from the vehicle number recognition unit (224) is transmitted to the server computer (400) through the portable communication device (230), and the monitored vehicle ( 100) Confirming whether the vehicle is a problem vehicle based on the vehicle number of 100) The driver or the passenger can confirm the result data regarding whether the problem occurs in the monitored vehicle (100) received from the server computer (400). Controls text and audio output.

  The vehicle number recognition unit (224) extracts the vehicle number of the monitored vehicle based on the imaging data of the monitored vehicle (100) through the imaging data input unit (222) under the control of the control unit (223).

  The database (225) is a vehicle number of the monitored vehicle (100) based on various programs necessary for the operation of the monitoring means (220) and the captured data of the monitored vehicle (100) captured by the camera (210). An automatic recognition program for the vehicle number is extracted for extraction.

  The display unit (226) displays, on the screen, result data regarding whether or not a problem occurs in the monitored vehicle (100) received from the server computer (400) through the portable communication device (230) under the control of the control unit (223). So that the driver or passenger can check.

  The voice output unit (227) converts the result data regarding whether or not a problem occurs in the monitored vehicle (100) received from the server computer (400) through the portable communication device (230) under the control of the control unit (223) into voice. Output through the rear speaker so that the driver or passenger can check.

  The communication interface (228) is communicatively connected to a portable communication device (230) provided in the monitoring vehicle (200), and the vehicle number of the monitored vehicle (100) is portable under the control of the control unit (223). Output to the communication device (230), and output to the control unit (223) the result data regarding whether or not a problem occurs in the monitored vehicle (100) input from the server computer (400) through the portable communication device (230).

  FIG. 3 is a diagram showing in more detail the configuration of the server computer 400 of FIG. 1 described above according to the first embodiment of the present invention.

  As shown in the figure, the data communication unit (410) is connected to the police (500), and various problems stolen and arranged from the computer (500) of the processing organization that controls and manages the vehicle, such as the police, the scrap car office, and the ward office. The vehicle information is periodically input, is output to the main control unit (420), and is monitored from the portable communication device (230) provided in the monitoring vehicle (200) through the communication network (300). ) Is received and output to the main control unit (420), and the result data for the problem occurrence of the monitored vehicle (100) input from the main control unit (420) is used as the monitoring vehicle (200). It transmits to the portable communication apparatus (230) with which it was equipped.

  The main control unit (420) receives various stolen vehicles arranged from the computer (500) of a processing organization that controls and manages vehicles such as police, scrap car offices, ward offices, etc. through the data communication unit (410). The vehicle number of the monitored vehicle (100) which is controlled to store the information in the database (430) and received from the portable communication device (230) provided in the monitored vehicle (200) through the data communication unit (410). Is used as a basis, the vehicle number of the monitored vehicle (100) is compared with the vehicle information stored in the database (430) to check whether a problem occurs in the monitored vehicle (100), and the monitored vehicle (100 ) Is transmitted to the portable communication device (230) provided in the monitoring vehicle (200).

  The database (430) updates and stores information on stolen and arranged various problem vehicles input from the processing engine computer (500) under the control of the main control unit (420).

  Next, a first embodiment of a method of automatically recognizing a monitored vehicle number using a monitoring vehicle according to the present invention configured as described above will be described in detail with reference to FIGS.

  FIG. 4 is a flowchart showing a first embodiment of a method for automatically recognizing a vehicle number of a monitored vehicle using the monitoring vehicle according to the present invention. FIGS. 5 to 8 are vehicle numbers of the monitored vehicle shown in FIG. 5 is a flowchart showing the automatic recognition routine in detail.

  First, in the first embodiment of the present invention, the vehicle number of the monitored vehicle (100) is extracted from the monitored vehicle (200) and transmitted to the server computer (400), and the server computer (400) is based on the extracted vehicle number. ) Confirms whether or not the problem of the monitored vehicle (100) occurs, and then determines whether the monitored vehicle (100) has a problem depending on whether or not the monitored vehicle (200) can be processed. To the computer (500) of the vehicle, the position is within the shooting range from the specific monitoring vehicle (200) equipped with automatic vehicle number recognition equipment (camera, monitoring means, portable communication device, etc.) The monitored vehicle (100) that is running or parked is recognized (S102), and the monitored vehicle is passed through the camera (210) provided in the monitored vehicle (200). Shooting, including the license plate information of 100) (S104).

  At this time, the monitored vehicle (100) is photographed by automatically recognizing the monitored vehicle (100) by a sensing sensor provided at a predetermined position of the monitored vehicle (200) or the camera (210), and then the camera (210). Or the driver or passenger can recognize the monitored vehicle (100) within the shooting range and operate the camera (210) to take a picture.

  After acquiring photographing data including license plate information of the monitored vehicle (100) through the camera (210), the control unit (223) of the sensing means (220) provided in the monitoring vehicle (200) captures the photographing data input unit. The shooting data input through (222) is output to the vehicle number recognition unit (224) to extract the vehicle number of the monitored vehicle (100) (S106).

  The vehicle number extraction process of the monitored vehicle 100 will be described in more detail with reference to FIGS. 5 to 8 of the accompanying drawings.

  First, in the vehicle number recognition unit (224), the license plate information of the monitored vehicle (100) photographed by the camera (210) from the photographing data of the monitored vehicle (100) input from the control unit (223). The license plate area of the monitored vehicle (100) is extracted based on the included shooting data (S106a).

  Referring to FIG. 6 in detail, a video for extracting a license plate area (for example, resolution 640 ×) from shooting data including license plate information of the monitored vehicle (100) taken by the camera (210). 480) (S106a-1), down-sampling (for example, resolution 640 × 240) with a reduced resolution of the video obtained from the step (S106a-1) is performed, and the image quality due to camera movement is obtained. In order to reduce the decrease (S106a-2), the downsampled video is selected from video pyramids with different spatial resolutions (for example, 640 × 240, 320 × 120, 160 × 60, 80 × 30) (S106a-3), and image pyramids with different spatial resolutions After that, the color information of the license plate is used to convert the hue to a gray level, and by selecting an appropriate threshold value, it is converted to a discretized image, generating a horizontal / vertical projection image and horizontal / vertical projection. Clustering is performed to extract a rectangular area using the aspect ratio of the image and the license plate (S106a-4), and the position of the license plate region from the clustered image If the position and size of the license plate area are incorrect from the clustered video, the next position (S106b-5) is determined. The control unit (223) of the monitoring means (220) provided in the monitoring vehicle (200) by generating information and a zoom control signal ) To control the camera (100) (S106a-6). That is, the step (S104) of photographing the monitored vehicle (100) through the camera (210) provided in the monitoring vehicle (200) (step S204 in the second embodiment described later, step S304 in the third embodiment). ) Is performed again.

  After extracting the license plate area of the monitored vehicle (100) in the above-described step (S106a), an individual character area is extracted from the extracted license plate area (S106b).

  Referring to FIG. 7 in detail, in the vehicle number recognition unit (224), the license plate area extracted in the above-described step (S106a) is converted into a gray level (S106b-1) and converted into a gray level. After applying the horizontal histogram to the license plate area and dividing it into an upper part and a lower part (S106b-2), the vertical histogram is applied to the license plate area divided into the upper part and the lower part according to the application of the horizontal histogram. Individual characters existing in the plate region are extracted (S106b-3), and the individual characters extracted through the horizontal / vertical histogram application in the above-described steps (S106b-2) and (S106b-3) are discrete through selection of appropriate threshold values. Then, the next stage (S106c) is performed (S106b-4).

  After the individual character area is extracted in the above-described step (S106b), a feature vector is extracted from each extracted individual character area (S106c).

  Describing in detail with reference to FIG. 8, after dividing each individual character of the license plate area extracted in the above-described step (S106b) in the vertical direction (S106c-1), it is divided in the horizontal direction (S106c-2). ) After dividing each individual character in the license plate area in the vertical / horizontal direction, the 64-dimensional (8 × 8) mapping is performed (S106c-3), and the 64-dimensional (8 × 8) mapping is performed. Normalizes the value of each mapped individual character and extracts a feature vector (S106c-4). At this time, since each mapped dimension has a value between 0 and 1, it is not necessary to normalize the character size.

  After the feature vector is extracted from the individual character area in the above-described step (S106c), the character and the number are recognized from the extracted feature vector, and the vehicle number of the monitored vehicle (100) is finally confirmed ( S106d). At this time, when recognizing characters and numbers based on the feature vector extracted from the individual character area, the character with the highest similarity is covered among the various characters recognized through pattern learning using a modular neural network. This is selected as the vehicle number of the monitoring vehicle (100).

  Then, after the vehicle number of the monitored vehicle (100) is extracted in the above-described step (S106), the monitoring means (220) uses the extracted vehicle number of the monitored vehicle (100) as the communication interface unit (228). Is transmitted to the server computer (400) through the portable communication device (230) connected to (S108).

  The server computer (400) receives the vehicle number of the monitored vehicle (100) from the monitored vehicle (200) through the communication network (300) (S110).

  Then, the main control unit (420) of the server computer (400) determines whether or not the information on the problem vehicle stored in the database (430) is the latest information within the valid period (S112).

  As a result of the determination, if the information on the problem vehicle stored in the database (430) is not the latest information within the valid period, the main control unit (420) can use the data communication unit (410) to police, scrap car office, ward office, etc. The communication connection is made to the computer (500) of the related institution, and the database (430) is updated by receiving the latest information (S114).

  As a result of the determination in the above-described step (S112), whether the information on the problem vehicle stored in the database (430) is the latest information within the effective period, or after updating the latest information through the above-described step (S114), the main information The control unit (420) compares the vehicle number of the monitored vehicle (100) received from the monitored vehicle (200) with the information on the problem vehicle stored in the database (430) (S116). It is determined whether the monitoring vehicle (100) is a problematic vehicle (S118).

  As a result of the determination, if there is no particular problem with the monitored vehicle (100), the main control unit (420) sends a message notifying that the monitored vehicle (100) is not a problem vehicle through the data communication unit (410). It transmits to the said monitoring vehicle (200) (S120).

  At this time, if the driver or passenger of the monitoring vehicle (200) knows that a request for vehicle inquiry has been made to the server computer (400), the monitoring means (220) of the monitoring vehicle (200) displays the display unit. (226) or the voice output unit (227) to notify that there is no problem with the monitored vehicle (100) so that the driver or passenger confirms it, and the driver or passenger of the monitored vehicle (200) If the person does not know that the vehicle inquiry request has been made, the monitoring means (220) of the monitoring vehicle (200) sends a message from the server computer (400) that there is no problem with the monitored vehicle (100). Record that it was received.

  However, if the monitored vehicle (100) is a problematic vehicle as a result of the determination in the above-described step (S118), the main control unit (420) processes the monitored vehicle (100) from the monitored vehicle (200). It is determined whether or not it is possible (S122). That is, it is confirmed whether the monitoring vehicle (220) is a vehicle having a judicial processing capability (for example, a police vehicle).

  As a result of the determination, if the monitored vehicle (100) having a problem can be processed directly from the monitored vehicle (200), the monitored vehicle (100) is a problem-occurring vehicle in the main control unit (420). A message informing that is transmitted (S124), and the monitored vehicle (100), which is a problem vehicle, can be processed in accordance with the processing procedure given in the monitored vehicle (200) (S126).

  However, if the monitored vehicle (200) does not have the judicial processing capability as a result of the determination in the above-described step (S122), if the problematic monitored vehicle (100) cannot be directly processed, the main control unit (420 ) Informs the computer (500) of the relevant processing organization such as the police (S128) of the information on the monitored vehicle (100) that is the problem vehicle (S128), and based on the information from the monitoring vehicle (200) in the processing organization. The problem monitored vehicle (100) can be processed (S130).

  Finally, the monitoring vehicle (200) or the vehicle on the related processing engine side records the shooting and processing results of the specific monitored vehicle (100), and the server computer (400) is also requested by the monitoring vehicle (200). Inquiry results and processing results of the monitored vehicle (100) are recorded and managed.

  The basic configuration of the second embodiment of the vehicle number automatic recognition system according to the present invention is the same as that of the first embodiment described above with reference to FIG. After extracting the vehicle number of the monitored vehicle (100) from the vehicle (200), the extracted vehicle number is compared with a database existing inside the monitored vehicle (200) to determine whether a problem occurs in the monitored vehicle (100) It has a functional difference that confirms immediately.

  That is, the monitoring vehicle (200) has a camera (210a) that captures the license plate information of the monitored vehicle (100) that is operating or parked, and the monitored vehicle that is based on the data captured through the camera (210a). The vehicle number of (100) is extracted, the vehicle number of the monitored vehicle (100) extracted is compared with the problem vehicle information stored in the database, and whether or not a problem occurs in the monitored vehicle (100) is confirmed. The result data on whether or not the problem of the monitored vehicle (100) can be reported to the driver or passenger by text or voice, but the server computer (400) is not notified of the result of whether or not the problem of the monitored vehicle (100) can occur. If possible, the monitoring means (220a) for recording the discovery of the problem vehicle and the information on the problem vehicle updated periodically from the server computer (400) If the server computer (400) can notify the server computer (400) of the result of whether or not a problem occurs in the monitored vehicle (100), the information on the monitored vehicle (100) that has been confirmed as the problematic vehicle is communicated. And a portable communication device (230a) for transmitting to the server computer (400) through the network (300).

  At this time, the monitoring means (220a) provided in the monitoring vehicle (200) described above is a theft stored in a database based on the vehicle number of the monitored vehicle (100) taken and extracted by the camera (210a), When comparing the information of various arranged problem vehicles and confirming whether or not a problem has occurred, if the monitored vehicle (100) is confirmed as a problem vehicle, whether or not the process in the monitored vehicle (200) is possible. If it is impossible to process the monitored vehicle (100) identified as a problem vehicle from the monitored vehicle (200), the data stored in the database is used as a basis and linked to the monitoring means (220a). Information on the monitored vehicle (100) that has been confirmed as a problem vehicle to the computer (500) of the processing organization such as the police, scrap car office, ward office through the portable communication device (230a) Both numbers, so as to process and transmit the current position, etc.).

  The server computer (400) obtains information on various problem vehicles that are periodically stolen and arranged from the processing organization computer (500) that controls and manages vehicles such as police, scrap car offices, and ward offices through the communication network (300). The information on the problem vehicles that are provided in the database and periodically updated is transmitted to the portable communication device (230a) provided in each vehicle (200) through the communication network (300), and the communication network (300 ), Information on the problem vehicle received from the portable communication device (230a) provided in the monitoring vehicle (200) and the monitored vehicle (100) confirmed is stored in the database.

  FIG. 9 is a view showing in more detail the internal structure of the monitoring vehicle 200 of FIG. 1 according to the second embodiment of the present invention.

  As shown in the figure, the key operation unit (221a) receives the key operation signal of the vehicle driver or passenger for recognizing the vehicle number of the monitored vehicle (100) and outputs it to the control unit (223a).

  The photographing data input unit (222a) receives photographing data of the monitored vehicle (100) photographed through the camera (210a) and outputs it to the control unit (223a).

  The control unit (223a) controls to store in the database (225a) information on theft and arranged various problem vehicles periodically received from the server computer (400), and through the shooting data input unit (222a). Based on the input photographing data of the monitored vehicle (100), the vehicle number recognition unit (224a) controls the vehicle number of the monitored vehicle (100) to be extracted, and the vehicle number recognition unit (224a) passes through the control. The extracted vehicle number of the monitored vehicle (100) and the problem vehicle information stored in the database (225a) are compared to confirm whether the monitored vehicle (100) has a problem. Characters and voice output are controlled so that the driver or passengers can check the result data for whether or not a problem has occurred, but temporarily monitored by the server computer (400) If the notification to the problem occurs whether the two (100) is not possible, to store the discovery of problems vehicle database (225a).

  The vehicle number recognition unit (224a) determines the vehicle number of the monitored vehicle (100) based on the imaging data of the monitored vehicle (100) input through the imaging data input unit (222a) under the control of the control unit (223a). Extract.

  The database (225a) uses the various programs necessary for the operation of the monitoring means (220a) and the imaged data of the monitored vehicle (100) captured by the camera (210a) as a basis for the vehicle number of the monitored vehicle (100). A vehicle number automatic recognition program for extraction is stored, and information on various trouble vehicles arranged and stolen periodically received from the server computer (400) is updated and stored, and the control unit (223a ) Save problem vehicle discovery by control.

  At this time, in the above-described database (225a), a compact disc (CD), which is a convenient storage medium, can be used for information on various problematic vehicles to be updated and managed.

  The display unit (226a) controls the control unit (223a) to display the result data on whether or not a problem occurs in the monitored vehicle (100) on the screen so that the driver or the passenger can check.

  The voice output unit (227a) converts the result data regarding whether or not the monitored vehicle (100) has a problem to be output under the control of the control unit (223a), and then outputs the voice through the speaker so that the driver or the passenger can check. To do.

  The communication interface unit (228a) is communicatively connected to the portable communication device (230a) provided in the monitoring vehicle (200), and is periodically received from the server computer (400) through the portable communication device (230a). The information on various stolen and arranged problem vehicles is output to the control unit (223a), and the information on the monitored vehicle (100) confirmed as the problem vehicle by the control of the control unit (223a) is transmitted to the portable communication device ( 230a).

  FIG. 10 is a view showing in more detail the configuration of the server computer 400 of FIG. 1 described above according to the second embodiment of the present invention.

  As shown in the figure, the data communication unit (410a) receives theft periodically input from a computer (500) of a processing institution that controls and manages vehicles such as police, scrap car offices, and ward offices through a communication network (300). The information of various arranged problem vehicles is transmitted to the portable communication device (230a) provided in the monitoring vehicle (200) through the communication network (300), and is provided in the monitoring vehicle (200) through the communication network (300). Information on the monitored vehicle (100) identified as the problem vehicle is received from the portable communication device (230a).

  The main control unit (420a) is arranged through the data communication unit (410a) to be periodically stolen and arranged from the computer (500) of the processing organization that controls and manages the vehicle, such as police, scrap car office, ward office, etc. The data communication unit is controlled to store information on various problem vehicles in the database (430a) and to be transmitted to the portable communication device (230a) included in the monitoring vehicle (200) through the communication network (300). Control is performed so that the information on the problem vehicle and the monitored vehicle (100) confirmed from the portable communication device (230a) provided in the monitoring vehicle (200) through (410a) is stored in the database (430a).

  The database (430a) updates and stores information on stolen and arranged various problem vehicles input from the computer (500) of the processing engine under the control of the main control unit (420a), and prepares for the monitoring vehicle (200). The information of the monitored vehicle (100) confirmed as the problem vehicle received from the portable communication device (230a) is stored.

  Next, a second embodiment of a method for automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle according to the present invention configured as described above will be described in detail with reference to FIG.

  FIG. 11 is a flowchart showing a second embodiment of a method for automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle according to the present invention.

  First, in the second embodiment of the present invention, after extracting the vehicle number of the monitored vehicle (100) from the monitored vehicle (200), the extracted vehicle number is compared with the database in the monitored vehicle (200). Confirm whether or not a problem occurs in the monitored vehicle (100), and immediately notify the driver / passenger of information on the monitored vehicle (100) in which the problem occurs in accordance with whether or not the monitored vehicle (200) can be processed. Or a specific monitoring vehicle (200) equipped with a vehicle number automatic recognition device (camera, monitoring means, portable communication device, etc.) for reporting to the computer (500) of the related processing organization for processing. The monitored vehicle (100) that is in operation or parked within the shooting range is recognized (S202) and the monitored vehicle (200a) provided in the monitored vehicle (200) is monitored ( 00) shooting, including the license plate information of (S204).

  At this time, the monitored vehicle (100) is photographed by the camera (210a) after the monitored vehicle (100) is automatically recognized by a monitoring sensor provided at a predetermined position of the monitored vehicle (200) or the camera (210a). It is possible to use a method in which a driver or a passenger recognizes the monitored vehicle (100) and operates the camera (210a) to take a picture.

  After acquiring the shooting data including the license plate information of the monitored vehicle (100) through the camera (210a), the control unit (223a) of the monitoring means (220a) provided in the monitoring vehicle (200) The photographing data input through the input unit (222a) is output to the vehicle number recognition unit (224a), and the vehicle number of the monitored vehicle (100) is extracted (S206). At this time, the extraction of the vehicle number based on the imaging data of the monitored vehicle (100) performed by the vehicle number recognition unit (224a) is the same as the vehicle number extraction process described in the first embodiment. Detailed description here will be omitted.

  After extracting the vehicle number of the monitored vehicle (100) through the step (S206) described above, the problem vehicle stored in the database (225a) provided in the monitoring means (220a) in the monitoring means (220a). It is determined whether the information is the latest information within the valid period (S208).

  As a result of the determination, if the information on the problem vehicle stored in the database (225a) is not the latest information within the valid period, the control unit (223a) of the monitoring means (220a) is connected to the communication interface unit (228a). Requests the server computer (400) through the portable communication device (230a) for the latest information on various problem vehicles provided on the computer (500) side of the related processing organization such as police, scrap car office, ward office, etc. (S210). The latest information on various problem vehicles is received from (400) and the database (225a) is updated (S212).

  As a result of the determination in the step (S208) described above, the information on the problem vehicle stored in the database (225a) is the latest information within the valid period, or the latest information was updated in the step (S210, S212) described above. Thereafter, the monitoring means (220a) compares the vehicle number of the monitored vehicle (100) extracted in the above-described step (S206) with the information on the problem vehicle stored in the database (225a) (S214), and compares the comparison. As a result, it is determined whether or not there is a problem with the monitored vehicle (100) (S216).

  If there is no special problem in the monitored vehicle (100) as a result of the determination, the monitoring means (220a) displays a message notifying that the monitored vehicle (100) is not a problem vehicle in the display unit (226a) or voice. It outputs to a character or a voice through an output part (227a), and a driver | operator or a passenger can confirm now (S218).

  However, if the monitored vehicle (100) is a problem vehicle as a result of the determination in the above-described step (S216), whether or not the monitored vehicle (100) can be processed in the monitoring means (220a). Is determined (S220). That is, it is confirmed whether the monitoring vehicle (200) is a vehicle having judicial processing capability (for example, a police vehicle).

  As a result of the determination, if the monitored vehicle (100) in which the problem is directly generated can be processed in the monitored vehicle (200), the monitoring means (220a) informs that the monitored vehicle (100) is a problem-occurring vehicle. Can be output as text or voice through the display unit (226a) or the voice output unit (227a) (S222), and the monitored vehicle (100), which is a problem vehicle, can be processed by the processing procedure given in the monitoring vehicle (200). (S224).

  However, if the monitoring vehicle (200) does not have judicial processing capability and cannot directly process the problematic monitored vehicle (100) as a result of the determination in the above-described step (S220), the monitoring means (220a) Then, through the portable communication device (230a) connected to the communication interface unit (228a), information on the monitored vehicle (100), which is a problem vehicle, is sent to the computer (500) of the related processing engine through the portable communication device (230a). (S226), the processing engine can process the monitored vehicle (100) having a problem based on the information from the monitored vehicle (200) (S228).

  Finally, in the monitoring vehicle (200) or the vehicle on the related processing engine side, the photographing and processing results of the specific monitored vehicle (100) are recorded, and the server computer (400) also requests from the monitoring vehicle (200). The inquiry result and the processing result of the monitored vehicle (100) are recorded and managed.

  The basic configuration of the third embodiment of the vehicle number automatic recognition system according to the present invention is the same as the configuration of FIG. 1 described from the first embodiment, although the reference numerals and names are the same. Then, all data obtained by photographing the monitored vehicle (100) from the monitored vehicle (200) is transmitted to the server computer (400), and the vehicle number of the monitored vehicle (100) is obtained from the server computer (400) based on the photographed data. After the extraction, the extracted vehicle number is compared with a database in the server computer (400) to confirm whether or not a problem occurs in the monitored vehicle (100).

  That is, the monitoring vehicle (200) has a camera number (210b) that captures the license plate information of the monitored vehicle (100) that is in operation or parked, and the vehicle number based on the imaging data of the monitored vehicle (100). After the extraction, the server computer (400) confirming whether or not the vehicle is a problem vehicle based on the extracted vehicle number, results data on whether or not a problem occurs in the monitored vehicle (100) is transmitted to the portable communication device (230b). Monitoring means (220b) that receives the data through the camera (210b) and transmits the data photographed through the camera (210b) to the server computer (400) through the communication network (300). And a device (230b).

  At this time, in the third embodiment of the present invention, as described above, the license plate of the monitored vehicle (100) that is operating or parked through the camera (210b) fixed to the monitoring vehicle (200). In addition to the method of capturing information, the license plate information of the monitored vehicle (100) being operated or parked by the driver / passenger of the monitoring vehicle (200) or the mobile (portable) camera holder is captured. After that, the mobile (portable) camera is connected to the monitoring means (220b), the captured data is downloaded, and the captured data including the license plate information of the monitored vehicle (100) is downloaded, A method of transmitting to the server computer (400) through the portable communication device (230b) connected to the monitoring means (220b) is also possible.

  The server computer (400) receives information on various problem vehicles that are stolen and arranged periodically from the processing engine computer (500) that controls and manages the vehicle, such as police, scrap car office, ward office, etc. through the communication network (300). To construct a database, extract the vehicle number of the monitored vehicle (100) based on the imaging data received from the monitored vehicle (200) through the communication network (300), and extract the monitored vehicle The vehicle number of (100) and the information on the problem vehicle stored in the database are compared to confirm whether the problem occurs in the monitored vehicle (100), and the result data for whether the problem occurs in the monitored vehicle (100) It transmits to the portable communication apparatus (230b) with which the said monitoring vehicle (200) was equipped, and the driver | operator or passenger of the monitoring vehicle (200) can confirm.

  At this time, the server computer (400) described above uses the imaging data of the monitored vehicle (100) received from the portable communication device (230b) included in the specific monitored vehicle (200) as a basis, 100) is extracted and compared with information on thefts and various problem vehicles arranged in the database, and whether the problem occurs is confirmed, the monitored vehicle (100) is confirmed as a problem vehicle. If it is determined whether or not the monitoring vehicle (200) is capable of processing, and if the monitored vehicle (100) that is confirmed as a problem vehicle is impossible in the monitoring vehicle (200), Information on the monitored vehicle (100) identified as a problem vehicle (vehicle number, current position, etc.) is transmitted to the processing computer (500) of the police, scrap car office, ward office, etc. for processing. To.

  FIG. 12 is a view showing in more detail the internal configuration of the monitoring vehicle 200 of FIG. 1 according to the third embodiment of the present invention.

  As shown in the figure, the key operation unit (221b) receives the key operation signal of the monitored vehicle driver or passenger for recognizing the vehicle number of the monitored vehicle (100) and outputs it to the control unit (223b). .

  The photographing data input unit (222b) receives photographing data of the monitored vehicle (100) photographed through the camera (210b) and outputs it to the control unit (223b).

  The control unit (223b) controls the imaging data of the monitored vehicle (100) input through the imaging data input unit (222b) to be transmitted to the server computer (400) through the portable communication device (230b), Characters and voice output are controlled so that the driver or passenger can check the result data regarding whether or not a problem occurs in the monitored vehicle (100) received from the server computer (400).

  The display unit (224b) displays, on the screen, result data regarding whether or not a problem occurs in the monitored vehicle (100) received from the server computer (400) through the portable communication device (230b) under the control of the control unit (223b). Check with the driver or passenger.

  The voice output unit (225b) uses the control unit (223b) to control the result data and data regarding whether or not a problem occurs in the monitored vehicle (100) received from the server computer (400) through the portable communication device (230b). After switching to, it is output through a speaker so that the driver or passenger can check.

  The communication interface unit (226b) is communicatively connected to a portable communication device (230b) provided in the monitoring vehicle (200) and is monitored by the camera (210b) under the control of the control unit (223b). The image data of (100) is output to the portable communication device (230b), and the result data for whether or not a problem occurs in the monitored vehicle (100) input from the server computer (400) through the portable communication device (230b) is controlled. Part (223b).

  FIG. 13 is a diagram showing in more detail the configuration of the server computer 400 shown in FIG. 1 according to the third embodiment of the present invention.

  As shown in the figure, the data communication unit (410b) is connected to the police, scrap car office, ward office, etc. through the communication network (300), and various vehicles stolen and arranged from the computer (500) of the processing organization that controls and manages the vehicle. Of the monitored vehicle (100) captured by the camera (210b) from the portable communication device (230b) provided in the monitored vehicle (200) through the communication network (300). The portable communication device (230b) included in the monitoring vehicle (200) is used to receive the result data for the occurrence of a problem in the monitored vehicle (100), which is received from the imaging data and input from the main control unit (420b). Transmit to.

  The main control unit (420b) is theft, various trouble vehicles arranged from the computer (500) of the processing organization that controls and manages the vehicle through the data communication unit (410b), such as police, scrap car office, ward office, etc. Is stored in the database (430b), and the image of the monitored vehicle (100) received from the monitored vehicle (200) through the data communication unit (410b) is used as a basis for the monitored vehicle (100 ), The vehicle number of the monitored vehicle (100) extracted is compared, and the vehicle number of the monitored vehicle (100) is compared with the vehicle information stored in the database (430b). Then, whether or not a problem occurs in the monitored vehicle (100) is confirmed, and result data on whether or not a problem occurs in the monitored vehicle (100) is provided in the monitored vehicle (200). It was controlled to be transmitted to the portable communication device (230b).

  The database (430b) updates and stores information on stolen and arranged various problem vehicles input from the computer (500) of each processing engine according to the control of the main control unit (420b).

  The vehicle number recognizing unit (440b) is configured to monitor the monitored vehicle (100) based on the captured data of the monitored vehicle (100) received from the monitored vehicle (200) through the data communication unit (410b) under the control of the main control unit (420b). 100) vehicle number is extracted.

  Next, a third embodiment of a method for automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle according to the present invention configured as described above will be described in detail with reference to FIG.

  FIG. 14 is a flowchart showing a third embodiment of a method for automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle according to the present invention.

  First, in the third embodiment of the present invention, in the server computer (400), after the vehicle number is extracted based on the imaged data of the monitored vehicle (100) received from the specific monitoring vehicle (200), whether or not a problem occurs In accordance with whether or not the monitoring vehicle (200) can be processed, information on the monitored vehicle (100) in which the problem has occurred is reported to the monitoring vehicle (200) or the computer (500) of the related processing organization. Operation or parking stop located within the shooting range from a specific monitoring vehicle (200) equipped with automatic vehicle number recognition equipment (camera, monitoring means, portable communication device, etc.) The monitored vehicle (100) is recognized (S302), and the license plate information of the monitored vehicle (100) is included through the camera (210b) provided in the monitored vehicle (200). Shooting (S304).

  At this time, the monitored vehicle (100) is photographed by automatically recognizing the monitored vehicle (100) by a sensing sensor provided at a predetermined position of the monitored vehicle (200) or the camera (210b), and then the camera (210b). Or a method in which the driver or a passenger recognizes the monitored vehicle (100) and operates the camera (210b) to shoot.

  After acquiring shooting data including license plate information of the monitored vehicle (100) through the camera (210b), the monitoring means (220b) of the monitoring vehicle (200) receives the shooting data of the monitored vehicle (100). Then, the data is transmitted to the server computer (400) through the portable communication device (230b) connected to the communication interface unit (226b) (S306).

  The server computer (400) receives imaging data of the monitored vehicle (100) from the monitored vehicle (200) through the communication network (300) (S308).

  Then, in the main control unit (420b) of the server computer (400), the image data of the monitored vehicle (100) received from the monitoring vehicle (200) through the data communication unit (410b) is used as the vehicle number recognition unit (440b). The vehicle number of the monitored vehicle (100) is extracted (S310). At this time, the extraction of the vehicle number based on the photographing data of the monitored vehicle (100) performed from the vehicle number recognition unit (440b) is the same as the vehicle number extraction process described in the first embodiment. Detailed description is omitted here.

  After the vehicle number is extracted by the vehicle number recognition unit (440b) based on the shooting data of the monitored vehicle (100), the main control unit (420b) uses the information on the problem vehicle stored in the database (430b). It is determined whether or not the latest information within the period (S312).

  If the information on the problem vehicle stored in the database (430b) is not the latest information within the valid period as a result of the determination, the main control unit (420b) uses the data communication unit (410b) to provide police, scrap car office, ward office, etc. The communication connection is made to the computer (500) of the related processing institution, and the latest information is inputted to update the database (430b) (S314).

  As a result of the determination in the step (S312) described above, the information on the problem vehicle stored in the database (430b) is the latest information within the effective period, or after the latest information is updated in the step (S314) described above, The main control unit (420b) compares the vehicle number of the monitored vehicle (100) extracted in the step (S310) described above with the information on the problem vehicle stored in the database (430b) (S316), and the comparison result. Then, it is determined whether or not there is a problem with the monitored vehicle (100) (S318).

  As a result of the determination, if there is no particular problem with the monitored vehicle (100), the main control unit (420b) sends a message informing that the monitored vehicle (100) is not a problem vehicle to the data communication unit (410b). Is transmitted to the monitoring vehicle (200) through (S320).

  At this time, if the driver or passenger of the monitoring vehicle (200) knows that the server computer (400) has requested vehicle inquiry, the monitoring means (220b) of the monitoring vehicle (200) displays The monitored vehicle (100) is informed that there is no problem through the unit (224b) or the voice output unit (225b) so that the driver or the passenger can check, while the driver of the monitored vehicle (200) Alternatively, if the passenger does not know that the vehicle inquiry request has been made, the monitoring means (220b) of the monitoring vehicle (200) says that there is no problem with the monitored vehicle (100) from the server computer (400). Record that the message has been received.

  However, if the monitored vehicle (100) is a problem vehicle as a result of the determination in the above-described step (S318), the main control unit (420b) processes the monitored vehicle (100) from the monitored vehicle (200). It is determined whether or not it is possible (S322). That is, it is confirmed whether the monitoring vehicle (200) is a vehicle having judicial processing capability (for example, a police vehicle).

  As a result of the determination, if the monitored vehicle (100) can be processed directly by the monitored vehicle (200), the main control unit (420b) sends the monitored vehicle (100) to the monitored vehicle (200) as a problem-occurring vehicle. A message notifying the existence is transmitted (S324), and the monitored vehicle (100) that is the problem vehicle can be processed according to the processing procedure given in the monitored vehicle (200) (S326).

  However, if the monitoring vehicle (200) does not have the judicial processing capability and cannot directly process the monitored vehicle (100) having a problem, the main control unit (420b) ) Informs the computer (500) of the relevant processing engine of the monitored vehicle (100) that is the problem vehicle (S328), and the processing engine determines the problem based on the information from the monitoring vehicle (200). A certain monitored vehicle (100) can be processed (S330).

  Finally, in the monitoring vehicle (200) or the vehicle on the related processing engine side, the photographing and processing results of the specific monitored vehicle (100) are recorded, and the server computer (400) also requests from the monitoring vehicle (200). The inquiry result and the processing result of the monitored vehicle (100) are recorded and managed.

  On the other hand, in the third embodiment of the present invention, the license plate information of the monitored vehicle (100) is obtained using the camera (210b) fixed to the monitoring vehicle (200) through the above-described step (S304). In addition to the shooting method, after the driver / passenger of the monitoring vehicle (200) or the licensed vehicle information of the monitored vehicle (100) being operated or parked by a mobile (portable) camera holder The mobile (portable) camera is connected to the monitoring means (220b) to download the photographed data, and the downloaded photographing data including the license plate information of the monitored vehicle (100) is downloaded to the monitoring means ( A method of transmitting to the server computer (400) through the portable communication device (230b) connected to 220b) can also be used.

  In the present invention as described above, by monitoring other vehicles through the monitoring vehicle and obtaining information, it is possible to reduce the cost put into the stolen vehicle and the arranged vehicle. Since it is possible to ascertain whether or not there is a problem with other vehicles that are directly monitored vehicles, it is effective for crime prevention.

  The present invention described above is not limited by the above-described embodiments and the accompanying drawings, and various substitutions, modifications and changes can be made without departing from the technical idea of the present invention. Will be clearly understood by those skilled in the art to which the present invention pertains.

The above object and other features of the present invention will become apparent from the following more detailed description of the preferred embodiment with reference to the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS The drawing which shows the structure of the automatic recognition system of the vehicle number of the to-be-monitored vehicle using the monitoring vehicle by each Example of this invention. 1 is a diagram showing a method for automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle according to a first embodiment of the present invention. 2 is a diagram showing in more detail the internal configuration of the server computer of FIG. 1 according to the first embodiment of the present invention. The flowchart with respect to the automatic recognition method of the vehicle number of the to-be-monitored vehicle using the monitoring vehicle by 1st Example of this invention. The flowchart which showed the automatic recognition routine of the vehicle number of the to-be-monitored vehicle of FIG. 4 in detail. The figure which showed further the internal structure of the monitoring vehicle and server computer by 2nd Example of this invention in detail. The flowchart which showed 2nd Example with respect to the automatic recognition method of the vehicle number of the to-be-monitored vehicle using the monitoring vehicle by this invention. The figure which showed further the internal structure of the monitoring vehicle and server computer by 3rd Example of this invention. The flowchart which showed 3rd Example with respect to the automatic recognition method of the vehicle number of the to-be-monitored vehicle using the monitoring vehicle by this invention.

Claims (16)

In the automatic recognition system for the vehicle number of a vehicle that is operating or parked using a surveillance vehicle,
A camera that captures the license plate information of a monitored vehicle that is operating or parked, and a vehicle number of the monitored vehicle is extracted based on data captured through the camera, and based on the extracted vehicle number of the monitored vehicle The monitoring means for receiving the result data on the possibility of the problem of the monitored vehicle from the server computer that confirms whether or not the vehicle is a problem vehicle, and notifying the driver or passenger by text or voice, and extracted from the monitoring means A monitoring vehicle equipped with portable communication equipment that transmits the vehicle number of the monitored vehicle to the server computer through the communication network; and a processing agency that controls and manages the vehicle, such as police, scrap car office, ward office, etc. through the communication network. Periodically from a computer, the information on various vehicles that were stolen and arranged was provided in the database, and through the communication network When the vehicle number of the monitored vehicle is received from the portable communication device provided in the monitored vehicle, the vehicle number of the monitored vehicle is compared with the vehicle information stored in the database, and a problem occurs in the monitored vehicle. Including a server computer that confirms the availability and transmits the result data for the occurrence of the problem of the monitored vehicle to the portable communication device provided in the monitored vehicle ,
The monitoring means includes
In order to extract an effective license plate area, after downsampling the video taken by the camera, the downsampled video is composed of video pyramids with different spatial resolutions, and the license plate color information is obtained. Use this to convert shades to gray levels, convert to binarized video by selecting the appropriate threshold value, use horizontal / vertical projection video generation, and use horizontal / vertical projection video and license plate horizontal / vertical ratio If the position and size of the license plate area are correct from the clustered video, the video is determined to be a valid license plate area and is not correct. The vehicle to be monitored by controlling the vertical and horizontal position information and zoom of the camera. Number plate information retake, extracted again a valid license plate area of
In order to extract characters from the valid license plate area, the extracted license plate area is converted to a gray level, and a horizontal histogram is applied to the license plate area converted to the gray level to divide it into an upper part and a lower part. Applying a vertical histogram to the license plate area divided into the upper part and the lower part to extract individual characters existing in the license plate area and extracting individual characters through binarization processing through appropriate threshold selection,
In order to identify the extracted individual characters, each individual character in the extracted license plate area is divided from each individual character in the vertical / horizontal direction, and then the 64-dimensional ( 8 × 8 ) mapping is performed, and then each individual individual mapped A system for automatically recognizing a vehicle number of a monitored vehicle using a monitored vehicle, wherein a character vector is normalized to extract a feature vector and an individual character is identified. In the first item, the monitoring means provided in the monitoring vehicle includes:
A key operation unit for receiving a key operation signal of a monitored vehicle driver or a passenger for recognizing the vehicle number of the monitored vehicle;
A photographing data input unit for inputting photographing data of the monitored vehicle photographed through the camera;
Based on the imaging data of the monitored vehicle input through the imaging data input unit, the extraction of the vehicle number of the monitored vehicle is controlled, and the extracted vehicle number of the monitored vehicle is transmitted to the server computer through the portable communication device. The driver or the passenger transmits the result data regarding whether or not the problem of the monitored vehicle is received, which is received from the server computer that confirms whether or not the vehicle is a problem vehicle based on the vehicle number of the monitored vehicle. Control unit that controls text and audio output so that it can be confirmed;
Based on the shooting data of the monitored vehicle input through the shooting data input unit under the control of the control unit, downsampling the shooting data, converting to gray level and clustering, and extracting an effective license plate area, A gray level conversion and horizontal / vertical histogram is applied to the extracted license plate area to extract individual characters, and each extracted individual character is divided in the vertical / horizontal direction to obtain 64 dimensions (8 × 8). A vehicle number recognition unit that performs mapping and then normalizes the value of each mapped individual character to extract a feature vector and identifies the individual character to extract the vehicle number of the monitored vehicle;
A database storing various programs necessary for the operation of the monitoring means and a vehicle number automatic recognition program for extracting the vehicle number of the monitored vehicle based on the imaged data of the monitored vehicle imaged by the camera ;
A display unit for displaying result data on whether or not a problem occurs in the monitored vehicle received from the server computer through the portable communication device under the control of the control unit;
A voice output unit that switches the result data regarding whether or not a problem occurs in the monitored vehicle received from the server computer through the portable communication device under the control of the control unit to a voice and outputs it through a speaker; and the portable communication device provided in the monitoring vehicle Result data on whether or not a problem occurs in the monitored vehicle, which is input from a server computer through the portable communication device by outputting the vehicle number of the monitored vehicle to the portable communication device under the control of the control unit And a communication interface unit for outputting the control number to the control unit. An automatic recognition system for a vehicle number of a monitored vehicle using a monitoring vehicle. In item 1, the server computer is:
Through the communication network, police, scrapped vehicle offices, ward offices, etc., such as processing and control of the vehicle, the computer of the processing organization that performs the control, management, the information on various trouble vehicles arranged and stolen is periodically input, and the monitoring through the communication network The mobile communication device provided in the monitored vehicle receives the vehicle number of the monitored vehicle from the portable communication device provided in the vehicle, and outputs the result data for the occurrence of the problem of the monitored vehicle input from the main control unit. Data communication part to be transmitted to equipment;
The data communication unit controls the database construction of information on theft and various troubled vehicles arranged from the computer of the processing organization that controls and manages the vehicle, such as police, scrap car office, ward office, etc. through the data communication unit, Based on the vehicle number of the monitored vehicle received from the portable communication device provided in the monitored vehicle through, the vehicle number of the monitored vehicle is compared with the vehicle information stored in the database. A main control unit that controls whether or not a problem occurs in the monitored vehicle and controls result data regarding whether or not the problem occurs in the monitored vehicle to be transmitted to a portable communication device provided in the monitored vehicle; and processing by the control of the main control unit It was configured including a theft input from the computer of the institution and a database that updates and stores information on various troubled vehicles arranged. An automatic recognition system for a vehicle number of a monitored vehicle using a monitoring vehicle. In the first or third item, the server computer is
Based on the vehicle number of the monitored vehicle received from the portable communication device provided in the specific monitoring vehicle, the information on theft and arrangement of various trouble vehicles arranged in the database is compared to confirm whether the problem occurs When the monitored vehicle is confirmed as a problem vehicle, it is determined whether or not processing by the monitored vehicle is possible,
If it is impossible to process a monitored vehicle that has been identified as a problem vehicle by the monitoring vehicle, information on the monitored vehicle that has been confirmed as a problem vehicle by the computer of a processing agency such as the police, scrap car office, ward office, etc. A system for automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle, wherein the vehicle number, current position, etc. of the vehicle are transmitted and processed. In the automatic recognition system for the vehicle number of a monitored vehicle that is operated or parked using a monitoring vehicle,
A camera that captures monitored license plate information that is in operation or parked, and a vehicle number of the monitored vehicle is extracted based on data captured through the camera, and the extracted vehicle number and database of the monitored vehicle are extracted. by comparing the problem vehicle information stored verify problem occurs whether the monitored vehicle, informs the driver or passenger of the result data for the problem occurrence whether the monitored vehicle characters, voice racemate, a server computer If it is impossible to notify the result of whether or not the problem of the monitored vehicle occurs, the monitoring means for recording the discovery of the problem vehicle and the information on the problem vehicle periodically updated from the server computer are received. If it is possible to notify the server computer of the result of whether or not there is a problem with the monitored vehicle, the information on the monitored vehicle that has been confirmed as the problem vehicle is supported via the communication network. Surveillance vehicles equipped with portable communication devices for transmission to the computer; and theft and arrangement periodically from computers of the processing agencies that control and manage the vehicles, such as police, scrap car offices, ward offices, etc., through the communication network Information on various problem vehicles is provided, a database is constructed, information on problem vehicles that are periodically updated is transmitted to the portable communication device provided in each monitoring vehicle through the communication network, and the monitoring vehicle is transmitted through the communication network. Including a server computer for storing in a database information on the problem vehicle and the monitored vehicle received from the portable communication device provided in the
The monitoring means includes
In order to extract an effective license plate area, after downsampling the video taken by the camera, the downsampled video is composed of video pyramids with different spatial resolutions, and the license plate color information is obtained. Use this to convert shades to gray levels, convert to binarized video by selecting the appropriate threshold value, use horizontal / vertical projection video generation, and use horizontal / vertical projection video and license plate horizontal / vertical ratio If the position and size of the license plate area are correct from the clustered video, the video is determined to be a valid license plate area and is not correct. The vehicle to be monitored by controlling the vertical and horizontal position information and zoom of the camera. Number plate information retake, extracted again a valid license plate area of
In order to extract characters from the valid license plate area, the extracted license plate area is converted to a gray level, and a horizontal histogram is applied to the license plate area converted to the gray level to divide it into an upper part and a lower part. Applying a vertical histogram to the license plate area divided into the upper part and the lower part to extract individual characters existing in the license plate area, extracting individual characters through binarization processing through appropriate threshold value selection,
In order to identify the extracted individual characters, each individual character in the extracted license plate area is divided from each individual character in the vertical / horizontal direction, and then the 64-dimensional ( 8 × 8 ) mapping is performed, and then each individual individual mapped A system for automatically recognizing a vehicle number of a monitored vehicle using a monitored vehicle, wherein a character vector is normalized to extract a feature vector and an individual character is identified. In the fifth aspect, the monitoring means provided in the monitoring vehicle includes:
A key operation unit to which a key operation signal of a monitored vehicle driver or a passenger for recognizing the vehicle number of the monitored vehicle is input;
A photographing data input unit for inputting photographing data of the monitored vehicle photographed through the camera;
Controls the storage of information on various stolen vehicles that are periodically received from the server computer and arranged various problem vehicles, and based on the shooting data of the monitored vehicle input through the shooting data input unit, Controls the extraction of the vehicle number, compares the extracted vehicle number of the monitored vehicle with the problem vehicle information stored in the database, confirms whether the problem of the monitored vehicle occurs, and determines whether the problem of the monitored vehicle occurs Characters and voice output are controlled so that the driver or passenger can check the result data for the vehicle, but if it is impossible to notify the server computer of the result of the problem occurrence of the monitored vehicle, the problem vehicle is found. Control unit to be stored in the database;
Based on the shooting data of the monitored vehicle input through the shooting data input unit under the control of the control unit, downsampling, gray level conversion and clustering are performed on the shooting data, and an effective license plate area is extracted and extracted. A gray level conversion and a horizontal / vertical histogram are applied to the license plate area, and individual characters are extracted. Each extracted individual character is divided in the vertical / horizontal direction, and is mapped in 64 dimensions (8 × 8). A vehicle number recognition unit that normalizes the value of each mapped individual character, extracts a feature vector, and extracts the vehicle number of the monitored vehicle by identifying the individual character;
A vehicle number automatic recognition program for extracting a vehicle number of a monitored vehicle is stored based on various programs necessary for the operation of the monitoring means and shooting data of the monitored vehicle taken by the camera. A database for periodically storing the information of theft and arranged various problem vehicles received from the server computer and storing the information of the problem vehicles under the control of the control unit;
A display unit for displaying result data on whether a problem occurs in the monitored vehicle under the control of the control unit;
A voice output unit that switches the result data regarding whether or not a problem occurs in the monitored vehicle to voice by the control of the control unit and outputs the result through a speaker; and a communication connection with a portable communication device provided in the monitoring vehicle; Information on theft and arranged various problem vehicles periodically received from the server computer through the communication device is output to the control unit, and the information on the monitored vehicle confirmed as the problem vehicle by the control of the control unit is portable An automatic recognition system for a vehicle number of a monitored vehicle using a monitoring vehicle, comprising a communication interface unit for outputting to a communication device. In item 5, the server computer is
Information on theft and arranged various problem vehicles periodically input from a computer of a processing agency that controls and manages vehicles such as police, scrap car offices, ward offices, etc. through the communication network, and the monitoring vehicles through the communication network A data communication unit that transmits information to a portable communication device included in the vehicle and receives information on the monitored vehicle that is confirmed as a problem vehicle from the portable communication device included in the monitoring vehicle through the communication network;
Controls the construction of a database of information on various types of stolen and arranged problem vehicles that are periodically input from the computer of a processing organization that controls and manages vehicles such as police, scrap car offices, ward offices, etc. through the data communication unit And confirming that the problem vehicle has been received from the portable communication device provided in the monitoring vehicle through the data communication unit, and controlled to be transmitted to the portable communication device provided in the monitoring vehicle through the communication network. A main control unit that controls to store the information on the monitored vehicle in the database; and updates information on various stolen and arranged problem vehicles input from the computer of the processing engine under the control of the main control unit. A database that stores and stores information on the monitored vehicle identified as the problem vehicle received from the portable communication device provided in the monitored vehicle. A system for automatically recognizing a vehicle number of a monitored vehicle using vehicle monitoring. In the paragraph 5 or 6, the monitoring means provided in the monitoring vehicle includes:
Based on the vehicle number of the monitored vehicle photographed and extracted by the camera, the monitored vehicle is identified as the problematic vehicle when comparing theft information stored in the database and checking information on various troubled vehicles arranged. If it is confirmed, it is determined whether or not processing by the monitoring vehicle is possible,
If it is impossible to process a monitored vehicle identified as a problem vehicle from the monitored vehicle, based on the data stored in the database, the police, scrapped vehicle establishment, through a portable communication device connected to the monitoring means, An automatic recognition system for a vehicle number of a monitored vehicle using a monitoring vehicle, wherein information of the monitored vehicle identified as a problem vehicle is transmitted to a computer of a processing organization such as a ward office and processed. In the automatic recognition system for the vehicle number of a monitored vehicle that is operated or parked using a monitoring vehicle,
Whether the vehicle is a problem vehicle based on the extracted vehicle number after extracting the vehicle number based on the camera that captures the license plate information of the monitored vehicle that is in operation or parked, and the captured data of the monitored vehicle Monitoring means for notifying the driver or passenger of the result of whether or not a problem occurs in the monitored vehicle from the server computer for confirming whether or not the character or voice is received through a portable communication device, and data taken through the camera A monitoring vehicle provided with a portable communication device that transmits the information to a server computer through a communication network; and periodically from a computer of a processing organization that controls and manages the vehicle, such as police, scrap car office, ward office, etc., through the communication network Information on various vehicles that were stolen and arranged was constructed in a database and monitored via the communication network The vehicle number of the monitored vehicle is extracted based on the captured data received from the vehicle, and the problem of the monitored vehicle is generated by comparing the extracted vehicle place number of the monitored vehicle with the problem vehicle stored in the database. A server computer for confirming whether or not a problem occurs in a monitored vehicle and transmitting result data on whether or not a problem vehicle occurs to a portable communication device provided in the monitored vehicle so that the driver or passenger of the monitored vehicle can confirm Include
The server computer
In order to extract an effective license plate area, after downsampling the video taken by the camera, the downsampled video is composed of video pyramids with different spatial resolutions, and the license plate color information is obtained. Use this to convert shades to gray levels, convert to binarized video by selecting the appropriate threshold value, use horizontal / vertical projection video generation, and use horizontal / vertical projection video and license plate horizontal / vertical ratio If the position and size of the license plate area are correct from the clustered video, the video is determined to be a valid license plate area and is not correct. The vehicle to be monitored by controlling the vertical and horizontal position information and zoom of the camera. License plate region retake, extracted again a valid license plate area of
In order to extract characters from the valid license plate area, the extracted license plate area is converted to a gray level, and a horizontal histogram is applied to the license plate area converted to the gray level to divide it into an upper part and a lower part. Applying a vertical histogram to the license plate area divided into the upper and lower sections to extract individual characters existing in the license plate area and extracting individual characters through binarization processing by selecting appropriate threshold values ,
In order to identify the extracted individual characters, each individual character in the extracted license plate area is divided from each individual character in the vertical / horizontal direction, and then the 64-dimensional ( 8 × 8 ) mapping is performed, and then each individual individual mapped A system for automatically recognizing a vehicle number of a monitored vehicle using a monitored vehicle, wherein a character vector is normalized to extract a feature vector and an individual character is identified. In the ninth aspect, the monitoring means provided in the monitoring vehicle includes:
A key operation unit of a monitored vehicle driver or a passenger for recognizing a vehicle number of a monitored vehicle;
A photographing data input unit for inputting photographing data of the monitored vehicle photographed through the camera;
Control is performed so that shooting data of the monitored vehicle input through the shooting data input unit is transmitted to the server computer through the portable communication device, and result data regarding whether or not a problem occurs in the monitored vehicle received from the server computer. Control unit that controls text and voice output so that the driver or passenger can check;
A display unit that displays result data for whether or not a problem occurs in the monitored vehicle that is received from the server computer through the portable communication device under the control of the control unit;
A voice output unit for switching the result data regarding whether or not a problem occurs in the monitored vehicle received from the server computer through the portable communication device under the control of the control unit to a voice and outputting it through a speaker; Of the monitored vehicle that is connected to the network and outputs the shooting data of the monitored vehicle captured by the camera under the control of the control unit to the portable communication device and is input from the server computer through the portable communication device A system for automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle, comprising: a communication interface unit for outputting result data on whether or not it is generated to the control unit. In item 9, the server computer is:
Through the communication network, police, scrapped vehicle offices, ward offices, etc., such as processing and control of the vehicle, the computer of the processing organization that performs the control, management, the information on various trouble vehicles arranged and stolen is periodically input, and the monitoring through the communication network Received from the portable communication device provided in the vehicle is captured data of the monitored vehicle photographed by the camera, and the monitored vehicle is provided with result data for whether or not a problem occurs in the monitored vehicle that is input from the main control unit. A communication unit for transmitting to a portable communication device provided;
The database communication unit controls the database construction of information on theft and various troubled vehicles arranged from the computer of the processing agency that controls and manages the vehicle, such as police, scrap car office, ward office, etc. through the data communication unit, The vehicle number of the monitored vehicle is controlled based on the vehicle number of the monitored vehicle received from the monitored vehicle through the vehicle number of the monitored vehicle based on the extracted vehicle number of the monitored vehicle The vehicle information stored in the database is compared to check whether or not a problem occurs in the monitored vehicle, and result data on whether or not a problem occurs in the monitored vehicle is transmitted to the portable communication device provided in the monitored vehicle. Main control unit to control as follows;
A database that updates and stores information on various stolen and arranged problem vehicles, which is input from a computer of a processing engine under the control of the main control unit; and
Based on the imaged data of the monitored vehicle received from the monitored vehicle through the database communication unit under the control of the main control unit, downsampling, gray level conversion and clustering are performed on the imaged data, and an effective license plate area is obtained. Extracting and applying the gray level conversion and the horizontal / vertical histogram to the extracted license plate area to extract individual characters, dividing each extracted individual character in the vertical / horizontal direction, 64 dimensions (8 × After the mapping of 8) is performed, a value of each mapped individual character is normalized, a feature vector is extracted, and a vehicle number recognition unit that extracts the vehicle number of the monitored vehicle by identifying the individual character is included. An automatic recognition system for a vehicle number of a monitored vehicle using a monitoring vehicle, characterized by being configured. In item 9 or 11, the server computer
The vehicle number of the monitored vehicle is extracted based on the photographed data of the monitored vehicle received from the portable communication device provided in the monitored vehicle, and information on thefts and various trouble vehicles arranged in the database is stored. In comparison, when confirming whether or not a problem has occurred, if the monitored vehicle is confirmed as a problem vehicle, it is determined whether or not processing by the monitored vehicle is possible,
If it is impossible to process a monitored vehicle that has been identified as a problem vehicle from the monitored vehicle, information on the monitored vehicle identified as the problem vehicle is transmitted to the computer of a processing agency such as the police, scrap car office, or ward office. A system for automatically recognizing a vehicle number of a monitored vehicle using a monitored vehicle, wherein In the ninth aspect, the monitoring means provided in the monitoring vehicle includes:
Downloaded when shooting data containing license plate information of a monitored vehicle being operated or parked is taken through a driver / passenger of a monitored vehicle or a mobile (portable) camera holder A method for automatically recognizing a vehicle number of a monitored vehicle using the monitored vehicle, wherein the data of the monitored vehicle is transmitted to the server computer through a communication network. In a method of automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle applied to an automatic vehicle number recognition system,
(1) a step of recognizing a monitored vehicle that is in operation or parked within a shooting range of a camera attached to the monitoring vehicle, and shooting through the camera including a license plate of the monitored vehicle;
(2) extracting the vehicle number of the monitored vehicle based on the shooting data of the monitored vehicle taken by the camera;
(3) transmitting the vehicle number of the monitored vehicle extracted from the monitoring means provided in the monitoring vehicle to the server computer through the portable communication device;
(4) When the vehicle number of the monitored vehicle is received from the monitored vehicle by the server computer, the latest information is obtained from the computer of the related processing engine according to the availability of the database provided for grasping the problem availability of the monitored vehicle. Receiving the information on the problem vehicle and updating the database;
(5) A step of confirming whether or not a problem has occurred by comparing information on the problem vehicle stored in the database based on the vehicle number of the monitored vehicle received from the monitored vehicle by the server computer;
(6) When it is confirmed that the monitored vehicle is not a problem vehicle, the server computer notifies the monitored vehicle that the monitored vehicle is not a problem vehicle;
(7) When the monitored vehicle is confirmed as a problem vehicle, the server computer determines whether the monitored vehicle can be processed from the monitored vehicle;
(8) If the monitored vehicle having the problem can be processed in the monitored vehicle, the server computer notifies the monitored vehicle that the monitored vehicle is a problem vehicle, and the problematic monitored vehicle Making it possible to process;
(9) In the monitoring vehicle, if the problem monitored vehicle cannot be processed, the server computer reports the monitored vehicle information to the computer of the related processing engine, and processes the problem monitored vehicle. Although made and a step for possible manner,
It said step (2) is
(2-1-1) obtaining an image for extracting a license plate area from shooting data including license plate information of a monitored vehicle shot by a camera;
(2-1-2) a step of downsampling to lower the resolution of the video acquired from the step (2-1-1) to mitigate a deterioration in image quality due to camera movement;
(2-1-3) In order to select a resolution suitable for extraction of the license plate region, the step of forming the downsampled video from video pyramids having different spatial resolutions;
(2-1-4) After being composed of video pyramids with different spatial resolutions, the hue information is converted to gray levels using the color information of the license plate, and converted to binary video by selecting the appropriate threshold value. Performing horizontal / vertical projection image generation and clustering to extract a rectangular region using the horizontal / vertical projection image and the aspect ratio of the license plate ;
(2-1-5) A step of determining whether the position and size of the license plate area are correct from the clustered video and performing the next step (2-2-1);
(2-1-6) When the position and size of the license plate area are not correct from the clustered video, the monitoring means provided in the monitoring vehicle by generating the vertical and horizontal position information of the camera and the zoom control signal To control the camera from the control unit
Step of converting the (6 2 - - 1) number plate area extracted by the gray level - (2-2-1) said step (2 1-1) to step;
(2-2-2) applying a horizontal histogram to the license plate region converted to the gray level to separate the license plate region into an upper part and a lower part;
(2-2-3) A step of extracting the individual characters existing in the license plate area by applying a vertical histogram to the license plate area separated into the upper and lower parts by applying the horizontal histogram;
(2-2-4) After binarizing the individual characters extracted by applying the horizontal / vertical histogram in the above steps (2-2-2) and (2-2-3) through selection of appropriate threshold values Performing the next step (2-3);
(2-3-1) a step of vertically dividing each individual character of the license plate area extracted by the steps (2-2-1) to (2-2-4);
(2-3-2) A step of dividing the individual characters in the license plate area in the vertical direction and then in the horizontal direction;
(2-3-3) Step of performing 64-dimensional (8 × 8) mapping after dividing each individual character in the license plate area in the vertical / horizontal direction; and
(2-3-4) A surveillance vehicle characterized in that it is made from the stage of normalizing the values of each mapped individual character and extracting a feature vector after mapping in 64 dimensions (8 × 8) A method for automatically recognizing a vehicle number of a monitored vehicle using a computer. In a method of automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle applied to an automatic vehicle number recognition system,
(A) A step of recognizing a monitored vehicle that is in operation or parked within a shooting range of a camera attached to the surveillance vehicle and photographs the vehicle including the license plate of the monitored vehicle through the camera;
(B) extracting the vehicle number of the monitored vehicle based on the imaging data of the monitored vehicle imaged by the camera;
(C) After extracting the vehicle number of the monitored vehicle, in the monitoring means provided in the monitored vehicle, it is related from the server computer according to the availability of the database provided to grasp the problem availability of the monitored vehicle Updating the database to receive the latest problem vehicle information provided to the computer of the processing agency;
(D) a step of confirming whether or not a problem has occurred by comparing information of the problem vehicle stored in the database based on the vehicle number of the monitored vehicle extracted from the monitoring means provided in the monitoring vehicle;
(E) When it is confirmed that the monitored vehicle is not a problem-occurring vehicle, the driver outputs a character or a voice notifying that the monitored vehicle is not a problem-occurring vehicle from the monitoring means provided in the monitored vehicle. Or to check with passengers;
(F) When the monitored vehicle is confirmed to be a problem vehicle, the monitoring means provided in the monitored vehicle determines whether the monitored vehicle can be processed;
(G) If the monitored vehicle having the problem can be processed in the monitored vehicle, the monitoring means provided in the monitored vehicle outputs a character or a sound notifying that the monitored vehicle is a problem vehicle. And allowing the driver or passenger to process the monitored vehicle in question; and (H) in the monitoring vehicle, if the processing of the monitored vehicle in question is not possible, Reporting the information of the monitored vehicle from the monitoring means provided to the computer of the relevant processing engine through the portable communication device, and processing the problematic monitored vehicle ,
In the step (B),
(B-1-1) obtaining an image for extracting a license plate area from shooting data including license plate information of a monitored vehicle shot by a camera;
(B-1-2) a step of down-sampling to lower the resolution of the video obtained from the step (B-1-1) to alleviate a deterioration in image quality due to camera movement;
(B-1-3) In order to select a resolution suitable for the extraction of the license plate area, the downsampled video is composed of video pyramids having different spatial resolutions;
(B-1-4) After being composed of video pyramids with different spatial resolutions, the color information of the license plate is used to convert the hue level to a gray level, and then converted to a binary video by selecting an appropriate threshold value. Performing horizontal / vertical projection image generation and clustering to extract a rectangular region using the horizontal / vertical projection image and the aspect ratio of the license plate ;
(B-1-5) determining whether the position and size of the license plate area are correct from the clustered video and performing the next step (B-2-1);
(B-1-6) When the position and size of the license plate area are not correct from the clustered video, the monitoring means provided in the monitoring vehicle by generating the position information and the zoom control signal of the camera up / down / left / right Controlling the camera from the controller of the camera;
(B-2-1) A step of converting the license plate area extracted by the steps (B-1-1) to ( B-1-6 ) to a gray level;
(B-2-2) applying a horizontal histogram to the license plate area converted to the gray level to separate the license plate area into an upper part and a lower part;
(B-2-3) Extracting individual characters existing in the license plate area by applying a vertical histogram to the license plate area separated into the upper part and the lower part by applying the horizontal histogram;
(B-2-4) After binarizing the individual characters extracted through the horizontal / vertical histogram application of the step (B-2-2) (B-2-3) by selecting an appropriate threshold value, Performing the next step (B-3);
(B-3-1) a step of vertically dividing each individual character of the license plate area extracted by the steps (B-2-1) to ( B-2-4 ) ;
(B-3-2) A step of dividing the individual characters in the license plate area in the vertical direction and then in the horizontal direction;
(B-3-3) Step of performing 64-dimensional (8 × 8) mapping after dividing each individual character in the license plate area in the vertical / horizontal direction; and
(B-3-4) After performing 64-dimensional (8 × 8) mapping, the value of each mapped individual character is normalized, and a feature vector is extracted. A method for automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle. In a method of automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle applied to an automatic vehicle number recognition system,
(A) a step of recognizing a monitored vehicle that is in operation or parked within a shooting range of a camera attached to the monitored vehicle, and shoots through the camera including a license plate of the monitored vehicle;
(B) transmitting captured image data of a monitored vehicle imaged by a camera to a server computer through a communication network;
(C) a step of extracting the vehicle number of the monitored vehicle based on the imaged data when the server computer receives the imaged data of the monitored vehicle from the monitored vehicle;
(D) The step of updating the database by receiving the latest information on the problem vehicle from the computer of the related processing engine according to the availability of the database provided for grasping the problem of the monitored vehicle in the server computer. ;
(E) A step of comparing the information on the problem vehicle stored in the database based on the vehicle number of the monitored vehicle from the server computer and confirming whether the problem occurs;
(F) When it is confirmed that the monitored vehicle is not a problem vehicle, the server computer notifies the monitored vehicle that the monitored vehicle is not a problematic vehicle;
(G) If the monitored vehicle is confirmed to be a problem vehicle, the server computer determines whether the monitored vehicle can process the monitored vehicle;
(H) In the monitored vehicle, if the problem monitored vehicle can be processed, the server computer notifies the monitored vehicle that the monitored vehicle is a problem vehicle, and the problematic monitored vehicle is identified. And (i) if it is impossible for the monitored vehicle to process the monitored vehicle in question, the server computer reports the monitored vehicle information to the computer of the related processing engine. Include steps to enable handling of problem monitored vehicles ,
Said step (c) comprises:
(C-1-1) obtaining an image for extracting a license plate area from shooting data including license plate information of a monitored vehicle shot by a camera;
(C-1-2) a step of down-sampling to lower the resolution of the video obtained from the step (c-1-1) to mitigate the deterioration of the image quality due to the movement of the camera;
(C-1-3) forming a downsampled video from video pyramids having different spatial resolutions in order to select a resolution suitable for the extraction of the license plate region;
(C-1-4) After being composed of video pyramids with different spatial resolutions, the color information of the license plate is used to convert the hue level to a gray level, and then converted to binary video by selecting an appropriate threshold value. Performing horizontal / vertical projection image generation and clustering to extract a rectangular region using the horizontal / vertical projection image and the horizontal / vertical ratio of the license plate;
(C-1-5) determining whether the position and size of the license plate area are correct from the clustered image and performing the next step (c-2-1);
(C-1-6) If the position and size of the license plate area are not correct from the clustered video, the monitoring means provided in the monitoring vehicle by generating positional information and zoom control signals of the upper, lower, left and right of the camera Controlling the camera from the controller of the camera;
(C-2-1) converting the license plate area extracted by the steps (c-1-1) to (c-1-6) into a gray level;
(C-2-2) applying a horizontal histogram to the license plate region converted to the gray level to separate the license plate region into an upper part and a lower part;
(C-2-3) A license plate area obtained by applying a vertical histogram to a license plate area separated into an upper part and a lower part by application of a horizontal histogram. Extracting individual characters present in
(C-2-4) After binarizing the individual characters extracted through the horizontal / vertical histogram application of the step (c-2-2) (c-2-3) by selecting an appropriate threshold value, Performing the next step (c-3-1);
(C-3-1) vertically dividing each individual character in the license plate area extracted by the steps (c-2-1) to (c-2-4);
(C-3-2) A step of dividing the individual characters in the license plate region in the vertical direction and then in the horizontal direction;
(C-3-3) performing a 64-dimensional (8 × 8) mapping after dividing each individual character in the license plate area in the vertical / horizontal direction; and
(C-3-4) After performing 64-dimensional (8 × 8) mapping, normalizing the mapped individual character values and extracting a feature vector, A method for automatically recognizing a vehicle number of a monitored vehicle using a monitoring vehicle.

Images